An extremely large cube of data may be distributed across a multi-node grid of connected computing devices due to the extremely large amount of disk space needed to store the cube of data. Multiple users may access the cube of data at any given time. It may be impractical however, to store multiple copies of the cube due to the extremely large amount of disk space needed as well as the memory needed to load the cube of data for use by users. Memory mapping technology may be used so that, as multiple users access the same cube, the cube is loaded into memory only once. In some cases, however, access is controlled to portions of the cube of data, for example, based on a user's position in a corporate structure, resulting in a need to create multiple copies of the cube to control the access.